Anomaly Detection and Localisation using Mixed Graphical Models

نویسندگان

  • Romain Laby
  • François Roueff
  • Alexandre Gramfort
چکیده

We propose a method that performs anomaly detection and localisation within heterogeneous data using a pairwise undirected mixed graphical model. The data are a mixture of categorical and quantitative variables, and the model is learned over a dataset that is supposed not to contain any anomaly. We then use the model over temporal data, potentially a data stream, using a version of the two-sided CUSUM algorithm. The proposed decision statistic is based on a conditional likelihood ratio computed for each variable given the others. Our results show that this function allows to detect anomalies variable by variable, and thus to localise the variables involved in the anomalies more precisely than univariate methods based on simple marginals.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

راهکار ترکیبی نوین جهت تشخیص نفوذ در شبکه‌های کامپیوتری با استفاده از الگوریتم-های هوش محاسباتی

In this paper, a novel hybrid method is proposed for intrusion detection in computer networks using combination of misuse-based and anomaly-based detection models with the aim of performance improvement. In the proposed hybrid approach, a set of algorithms and models is employed. The selection of input features is performed using shuffled frog-leaping (SFL) algorithm. The misuse detection modul...

متن کامل

Generalized Statistical Methods for Unsupervised Minority Class Detection in Mixed Data Sets

Minority class detection is the problem of detecting the occurrence of rare key events differing from the majority of a data set. This paper considers the problem of unsupervised minority class detection for multidimensional data that are highly nongaussian, mixed (continuous and/or discrete), noisy, and nonlinearly related, such as occurs, for example, in fraud detection in typical financial d...

متن کامل

Anomaly Detection and Modeling of Trajectories

The recent boom in the availability and use of geolocation technologies has created a great need to understand datasets of trajectories. Moreover, trajectory data is collected in a wide range of different domains including: meteorology, zoology, and business. However, trajectories have several intrinsic attributes that make them difficult to analyze. First, their time-series nature makes applyi...

متن کامل

An Introduction to Probabilistic Graphical Models for Relational Data

We survey some of the recent work on probabilistic graphical models for relational data. The models that we describe are all based upon ’graphical models’ [12]. The models can capture statistical correlations among attributes within a single relational table, between attributes in different tables, and can capture certain structural properties, such as the expected size of a join between tables...

متن کامل

Detection of Mo geochemical anomaly in depth using a new scenario based on spectrum–area fractal analysis

Detection of deep and hidden mineralization using the surface geochemical data is a challenging subject in the mineral exploration. In this work, a novel scenario based on the spectrum–area fractal analysis (SAFA) and the principal component analysis (PCA) has been applied to distinguish and delineate the blind and deep Mo anomaly in the Dalli Cu–Au porphyry mineralization area. The Dalli miner...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2017